The money in this market is in subscriptions. Recurring revenue from just a small percentage of all possible users can pay for a lot of bikes and it's pretty reliable revenue as long as you don't alienate customers too much.
I've been a Mobike subscriber for a year now (in Berlin). The bikes are crap (i.e. cheap, in need of maintenance) but they are everywhere and I see it as saving on gym cost (hard work to ride one). But, if I need to get from A to B, I get one because I already payed for it. I probably payed them around 100 euros or so over the last year. I just renewed for 3 months for 24 euros. Every time I do this the price seems to change. This gets me unlimited 30 minute rides. After that they charge 1 euro per additional 30 minutes (tip, lock and unlock seems to work). I consider this good value. I don't have to worry about bike repairs, bike theft, or leaving my bike in a dodgy area. As soon as I lock it, it's not my problem.
I've done over 1200 km on mobikes in the last 12 months. Even in the winter when the weather is less nice, I still used them enough to make it worth the subscription.
The back of the envelope math is pretty interesting. If you have 10K users like me that's about 1M revenue. For a city like Berlin (4M inhabitants), you can probably do better. At 100 euro per bike, 1M buys you 10K bikes. I have no idea what these things cost per year but both the upfront investment to flood a city with bikes and setting up the support organization (maintenance mainly) are not that high. I'd say if they shoot for recurring revenue from around 20K users in a city like this, they ought to be very profitable. They have plenty of capacity to improve utilization as at this rate most bikes are not in use most of the time.
The main issues I've seen with competitors is that they think too small. Something like Mobike only makes sense if you can find one without effort. A lot of the failed startups in this space did not have enough bikes or too complex bikes (they break). Launching with a few hundred bikes is pointless. Everybody that did this in Berlin is gone (ofo, byke (with a y), lime, etc. Mobike did this right. They flooded the city with between 5K and 10K bikes. They are all over the place. I suspect the bulk price for these things is pretty OK. The main challenge is the growth strategy: they need recurring revenue.
I suspect Jump, which recently re-launched here, won't make it in Berlin because they don't do subscriptions, their operational zone is tiny, and they don't have enough bikes on the streets. The value proposition kind of sucks compared to Mobike.
Mobike looks like it might fail because their bikes are deteriorating and they seem to be in limbo now that their China based headquarters looks like they want to focus on just China. Apparently they are looking to sell off the European operations. I think it's an operation worth saving but it is going to take an investment to keep their existing user base and there's a bit of urgency because the experience is getting worse as the bikes get shittier.
The only way to mess this up is poor execution. There seems a lot of that in this business. But it's not an inherently bad business.